A Study Of The Age Inference System Based On Multiplex Methylation SNaPshot For Forensic Bloodstain Analysis

Objective:Age inference can provide important information about the contributors of biological evidence left at crime scenes.In mammals,70%to 80%of cytosine at Cp G(5’-cytosine-phosphate-guanine-3’)sites are methylated,which has been proved age-related.Although the age-related Cp Gs reported so far were mainly screened by the Infinium Human Methylation450 Beadchip(HM450),HM450 mainly covers Cp G sites inside Cp G islands(regions with a high frequency of Cp G sites)rather than Cp G sites outside Cp G islands.This makes it hard to design primers for specific Cp Gs.Nevertheless,the newly released Infinium Human Methylation EPIC Beadchip(HM850)contains about 850,000 Cp Gs,covering much more Cp Gs outside Cp G islands.The golden standard for detecting specific Cp G sites is pyrosequencing,but it does not allow multiplex assay or matching with the most commonly used forensic genotyping platform,namely the capillary electrophoresis.In contrast,the multiplex methylation SNaPshot technology can detect multiple Cp G sites simultaneously and it is compatible with the capillary electrophoresis platform.This project aimed to construct an efficient age inference system based on multiplex methylation SNaPshot for forensic bloodstain analysis in Chinese population using the age-related Cp G sites selected from the public HM850 data.Moreover,the age inference system would be validated in additional sets of blood,fresh bloodstains,and aged bloodstains,respectively.Methods:To identify age-related candidate Cp G sites,we analyzed a DNA methylation dataset from the GEO(GSE149318)generated from 90 blood samples aged 22-51 years old using the HM850 platform.Then,we built a multiplex methylation SNaPshot detection assay to evaluate the candidate Cp G markers in a training set of 115 blood samples aged 11-71 years old and built an age inference model by full subset regression.Then,the age inference model was validated in three testing sets of blood samples,fresh bloodstain samples,and aged bloodstain samples from another 30 healthy individuals age 11-68 years old.Results:Data processing and correlation analysis were performed and a total of 36 CpGs were selected(|R|>0.47,p<0.01).Next,8 of the 36 Cp Gs were further extracted using stepwise regression.We failed amplification conditions for 3 of the 8 Cp Gs,and therefore we selected the remaining 5 Cp Gs to build the multiplex methylation SNaPshot detection assay.The blood training samples of 115 healthy individuals were tested using the multiplex methylation SNaPshot detection assay,after which the 5 Cp Gs were narrowed down to 3 by full subset regression.Three age inference models were trained respectively for the male group,female group,and gender-neutral group.The model for the male group had an adjusted coefficient of determination(adj R~2)of 0.879,a root-mean-square error(RMSE)of 4.717 years,and a mean absolute deviation(MAD)of 3.671 years.The model for the female group had an adj R~2of 0.862,an RMSE of 5.881 years,and a MAD of 4.219 years.The model for the gender-neutral group had an adj R~2of 0.864,an RMSE of 5.330years,a MAD of 4.038 years.We finally selected the age inference model for the gender-neutral group to construct the age inference system.The age inference model was Y=27.370-35.556*A+71.837*B-8.436*C(Y represents the inferred age,A,B,and C represent methylation rate of cg10501210,cg16867657,and cg13108341quantified by the multiplex methylation SNaPshot detection assay,respectively).Validation tests were performed using blood,fresh bloodstain,and aged bloodstain from 30 individuals.It was found that MAD was 4.734,4.490,and 5.431 years in the testing samples of blood,fresh bloodstain,and aged bloodstain,respectively.There were no statistically significant differences between the MAD values of the three testing sets and the age inference model.The results showed that the age inference system was stable and accurate for blood,fresh bloodstain,and aged bloodstain analysis.Conclusion:A methylation age inference system composed of a multiplex methylation SNaPshot detection assay and an age inference model was successfully constructed for bloodstain analysis in Chinese population.The methylation age inference system included 3 Cp Gs,among which cg10501210(C1orf132)and cg16867657(ELOVL2)were previously reported age-related Cp G,and cg13108341(DNAH9)was a newly discovered age-related Cp G.The blood methylation age inference system was suitable for blood,fresh,and aged bloodstain analysis within the age range of 11 to 71 years(adj R~2=0.864,RMSE=5.330 years,MAD=4.038years).which could provide a new tool for crime scene investigation.